Feasibility and reproducibility of semi-automated longitudinal strain analysis: a comparative study with conventional manual strain analysis

Table 2 Comparisons between the semi-automatic and manual strain measurements

Variable	Auto_1,%	Manual,%	p1	r	p2	Bias (LOAs)	Auto_2,%	p3	r	p4	Bias (LOAs)
LVGLS	-21.0 ± 2.5	-20.8 ± 2.4	0.230	0.84	< 0.001	-0.1(-2.9,2.6)	-21.1 ± 2.4	0.450	0.85	< 0.001	-0.1(-2.6,2.7)
RVFWS	-26.4 ± 4.8	-31.3 ± 5.8	< 0.001	0.25	< 0.005	4.9(-8.1,17.9)	-26.5 ± 4.)	0.886	0.74	< 0.001	0.1(-6.7,6.8)
LASr	48.0 ± 10.0	37.6 ± 9.9	< 0.001	0.33	< 0.001	10.5(-12.3,33.3)	47.7 ± 10.2	0.467	0.88	< 0.001	0.3(-9.4,10.0)
LAScd	-31.8 ± 8.9	-22.2 ± 7.9	< 0.001	0.55	< 0.001	-9.6(-25.4,6.2)	-31.8 ± 8.7	0.920	0.80	< 0.001	-0.0(-10.7,10.8)
LASct	-16.3 ± 6.5	-15.4 ± 6.0	0.176	0.24	< 0.005	-0.9(-16.1,14.3)	-16.0 ± 7.0	0.506	0.75	< 0.001	-0.3(-9.4,8.8)

Abbreviations LASr Left atrial reservoir strain, LAScd Left atrial conduit strain, LASct Left atrial pump strain, LVGLS Left ventricular global longitudinal strain, RVFWS Right ventricular free wall strain, Manual Manual strain measurement, Auto_1 Semi-automatic strain measurement of the first cardiac cycle, Auto_2 Semi-automatic strain measurement of the other cardiac cycle, p1 Paired t-test comparing the semi-automatic and manual measurements, p2 Linear regression assessing the relation between the semi-automatic and manual measurements, p3 Paired t-test comparing the semi-automatic measurements of two cardiac cycles, p4 Linear regression assessing the consistency between the semi-automatic measurements of two different cardiac cycles, LOAs limits of agreement

ISSN: 1476-7120